Comments are presented on a formula developed by Enzo Lombardo "for the relation between the death rate above age a and the growth rate above age a (where a is an advanced age, above 55 or 60) on the one hand, and the expectation of life at age a on the other."  Estimates obtained by this method are compared with results obtained by the method of S. Horiuchi and A. J. Coale.  It is concluded that "the formula proposed by Horiuchi and [Coale] produces much more accurate estimates, with a maximum error of 0.03 years."  The author describes a way in which Lombardo's method "can be modified to give estimates of precision comparable to the Horiuchi and Coale estimates."  (summary in FRE, ITA)